Ratiometric Fluorescent Sensor Array as a Versatile Tool for Bacterial Pathogen Identification and Analysis

Abstract

Rapid and reliable identification of pathogenic microorganisms is of great importance for human and animal health. Most conventional approaches are time-consuming and require expensive reagents, sophisticated equipment, trained personnel, and special storage and handling conditions. Sensor arrays based on small molecules offer a chemically stable and cost-effective alternative. Here we present a ratiometric fluorescent sensor array based on the derivatives of 2-(4'- N, N-dimethylamino)-3-hydroxyflavone and investigate its ability to provide a dual-channel ratiometric response. We demonstrate that, by using discriminant analysis of the sensor array responses, it is possible to effectively distinguish between eight bacterial species and recognize their Gram status. Thus, multiple parameters can be derived from the same data set. Moreover, the predictive potential of this sensor array is discussed, and its ability to analyze unknown samples beyond the list of species used for the training matrix is demonstrated. The proposed sensor array and analysis strategies open new avenues for the development of advanced ratiometric sensors for multiparametric analysis.

Keywords: 3-hydroxyflavone; ESIPT; Gram status; chemical nose; discriminant analysis; multiparametric sensing; pattern analysis; predictive analysis.

Figure 1

(A) Main protolytic equilibria and photochemical cycle of the 3-hydroxyflavone derivatives in ground and excited states; (B) experimental fluorescence spectrum of DMAF upon interaction with S. aureus (black line) and its computational deconvolution into components demonstrating emissions of normal (N*), anionic (A*) and tautomeric (T*) forms; (C) structures of 3-hydroxyflavone derivatives used as ratiometric reporters in the sensor array.

Figure 2

Single-channel ratiometric response from the four sensor array components upon interaction with (A) Gram-positive and (B) Gram-negative bacteria. Each response signal is an average of 12 replicate measurements. (C) Canonical plot of the discriminant analysis of the sensor response using first two discriminants. Each class (bacterial species) contains 12 experimental points representing replicate measurements. Each experimental point is an average of 5 independent measurements. Ellipses represent 95% confidence areas.

Figure 3

Canonical plot of the single-channel ratiometric response from the four sensor components in the subspace of the first three discriminants.

Figure 4

Dual-channel ratiometric response (T*/N* channel – solid bars, A*/N* channel – dashed bars) from the four sensor array components upon interaction with (A) Gram-positive and (B) Gram-negative bacteria. Each response signal is an average of 12 replicate measurements. (C) Canonical plot of the discriminant analysis of the sensor response using first two discriminants. Each class (bacterial species) consists of 12 experimental points representing replicate measurements. Each experimental point is an average of 5 independent measurements. Ellipses represent 95% confidence areas.

Figure 5

Recognition of the Gram status of the bacterial species using the discriminant analysis of the sensor array dual-channel responses.

Figure 6

Discrimination of a single bacterial species against all other species using the discriminant analysis of the sensor array dual-channel responses. Results for the Gram-positive bacteria are presented in the top row, and for Gram-negative bacteria are presented in the bottom row.